DISCUSSION OF RESULTS OBTAINED. 987 
Vanilla and vanillin are often replaced by balsam Peru, storax, tolu, and gum ben- 
zoin. 
The melting and congealing points of the fats and of the free fatty acids are no 
indication whatever, as pure cocoa butter shows great variations. 
Legler! regards the cellwose determination as an uncertain means of 
detecting husks, because the results vary with the method used. 
DISCUSSION OF THE RESULTS OBTAINED. 
While the time at our disposal has not permitted as thorough an 
investigation in some details of the work as seemed desirable, sufficient 
data have been obtained to show the general character and extent of 
the adulteration of the cocoa preparations sold on our markets. 
The determination of the ash serves as a means of determining the 
extent of dilution with foreign materials (the nature of these materials 
being known) of organic or inorganic origin. The per cent of ash 
decreases with the addition of substances low in ash, as starch, flour, 
ete.; increases with the removal of fat, the addition of mineral matter 
for weight or color, or the addition of alkaline carbonates in the process 
of manufacture. The percentages of ash, fat, sugar, and starch must 
be considered with reference to each other in order to determine whether 
any one is abnormal. <Any addition of mineral inatter that could pos- 
sibly be of value for increasing the weight or color of the preparation 
would certainly render the ash conspicuously abnormal in amount or 
color. 
For the detection of the use of fixed alkalis in the process of manu- 
facture, the number of cubic centimeters of decinormal acid necessary 
to neutralize the ash from 2 grams of material was determined. The 
number thus obtained is put down in the table of results as acid equiva- 
lent. By use of this number and the number expressing the per cent 
of ash in the sample, the following additional numbers were calculated : 
A = acid equivalent — per cent ash. 
B= acid equivalent — per cent ash — the per cent of alkali in the ash (calculated 
acid equivalent x .0053 x ) 
5) 
In the samples yielding a normal ash, A varies from 1.20 to 1.72 and 
exceeds 1.50 in eleven cases; B varies from 1.67 to 3.15 and exceeds 
2.80 in three cases. 
The determinations proposed by Stutzer were made in only six 
samples. It is greatly regretted that time did permit more of these 
determinations, since the method seems to be of considerable value, 
although the differences in the results are not. as marked as are those 
obtained by Stutzer (see pp. 945 and 985). 
The only decidedly abnormal ashes were those from samples Nos. 40, 
42, and 64 which were unquestionably prepared by use of alkalis (see 
table of results). The ash figures obtained for samples 51 and 53, while 
as Na,CO;) = acid equivalent — per cent ash — ( 
‘Rep. f- anal. chem., 1884, 4, 345, 
